mt-en-et-general-2

This model is a fine-tuned version of Helsinki-NLP/opus-mt-en-mul on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3021
• Bleu: 27.0505
• Gen Len: 24.461

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 128
• eval_batch_size: 128
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
0.4404	1.0	3236	0.3451	24.8436	24.71
0.3534	2.0	6472	0.3269	25.4530	24.4965
0.3284	3.0	9708	0.3178	26.1743	24.5935
0.3129	4.0	12944	0.3124	26.0742	24.5875
0.3022	5.0	16180	0.3090	26.5071	24.613
0.2941	6.0	19416	0.3065	26.3031	24.4815
0.2882	7.0	22652	0.3041	26.6162	24.4735
0.2836	8.0	25888	0.3033	26.6335	24.562
0.2804	9.0	29124	0.3024	26.9291	24.4725
0.2785	10.0	32360	0.3021	27.0505	24.461

Framework versions

• Transformers 4.57.3
• Pytorch 2.9.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.2